EP2394875A2 - Method for controlling the operation of at least one vicinity-sensing sensor system and motor vehicle - Google Patents

Abstract

The method involves delivering part of input data of a function, which evaluates the data for concrete control actions of a motorvheicle, by a sensor system. The sensor system is switched between two operating modes based on a determined control parameter e.g. parameter displaying unavoidableness of collision. A scanning frequency of a sensor e.g. infrared light sensor, an illumination area of the sensor system, illumination intensity, a sensor detection range, a number of objects to be evaluated, and evaluation accuracy are adapted based on the control parameter. An independent claim is also included for a motorvehicle comprising a sensor system.

Description

The invention relates to a method for controlling the operation of at least one sensor system encompassing the surrounding area, comprising at least one sensor and / or at least one evaluation module of a motor vehicle, which sensor system evaluates at least part of the input data of an input data for concrete control actions of the motor vehicle evaluating function of a vehicle system supplies, as well as an associated motor vehicle.

Today's sensor systems, in particular peripheral sensor systems, in addition to the at least one sensor, for example an environment sensor, often also an evaluation module, which is designed for processing the raw data of the sensor for the vehicle systems using them in the sense of a preliminary evaluation. Such an evaluation module can be realized, for example, as an electronic component, in particular as a processor, close to the sensor. For an environmental sensor, it may be provided, for example, that its raw data is already internally processed into an object list which may contain various detected objects and various properties of these objects in the form of a list. Of course, the processing can also include the derivation of further information from the raw data, for example in the sense of an interpretive processing of the raw data.

These data prepared in this way can be passed on to other vehicle systems, for example via a bus system such as a CAN bus, where they can be better processed by the individual functions. In the present case, a function is understood to mean any content-related functionality, the input data, in particular also of sensor systems, evaluates and generates therefrom control actions or control commands, which are transmitted to other vehicle systems. For example, a longitudinal guidance system has a distance-keeping function that deduces from input data about a preceding vehicle and speed information, for example, a control command to the brake system or the engine to keep the distance. A function can be realized by electronic and / or software components, for example a program package.

This sensor systems are usually operated with a set set of operating and evaluation parameters, which are not always adapted to the specific needs of each situation. In particular, with regard to the frequency or rate at which the sensor systems supply data, optimizations are often made on the entire range of functions, which in some cases, for example in critical situations, do not allow optimal performance.

This is explained in more detail using the example of a predictive safety system whose main function is collision monitoring. In this case, objects in the vicinity of the motor vehicle are detected by sensor systems and, in particular, the positions and relative speeds as well as information about the reaction possibilities of the motor vehicle are taken into account, for example to determine a collision probability and to take into account the activation of different safety systems of the motor vehicle. Particularly important is the behavior in an unavoidable collision, for example, if a time to collision (TTC) is determined, which in turn determines the triggering of security systems such as airbags or the like.

The data rates of the environment sensors, which is designed to capture large areas around the motor vehicle, are not designed for very short time windows, as they would be optimal for an upcoming collision for the so-called crash detection. In particular, there are usually many Data must be transmitted, the data transfer rate low. Thus, important information in the event of a collision can not be evaluated with the desired accuracy and frequency, since the sensor systems are not designed for crash detection.

The invention is therefore based on the object to provide a method with which a better overall utilization of sensor systems in motor vehicles is made possible and the aforementioned disadvantages are eliminated.

To solve this problem, it is provided according to the invention in a method of the type mentioned that the sensor system is switched between at least two operating modes used in the function depending on at least one control parameter determined within the framework of the function.

The invention therefore proposes, in the context of a single function of a vehicle system, for example a predictive safety system or a driver assistance system, to provide different types of operation of at least one sensor system used by the corresponding function. In principle, an operating mode can be understood as a predefined set of at least one operating parameter of the sensor and / or at least one evaluation parameter and / or at least one operating parameter of a device associated with the sensor, for example an illumination device. The control parameter, in turn, will generally reflect a particular aspect of the current situation of the motor vehicle which will cause the input data of the sensor system to be more useful to the function in modified form, such as more frequent and / or differently conditioned, and allow for improved performance of the function. In other words, the control parameter thus provides a possibility for internally relevant situation differentiation with regard to the function.

In this way, it is thus possible to operate sensor systems in such a way that the input data supplied to the function improves the requirements satisfy a high quality performance of the function; the possibilities of the sensor or sensor system are therefore utilized to a greater extent in the context of a function. The sensor system is thus adaptively adapted so that an optimized use of resources is made possible.

It should already be noted at this point that the operating modes can be predefined, that is, for example, different fixedly defined parameter sets for different values and / or value ranges of the at least one control parameter can be stored in a memory device. Of course, however, it can also be provided that at least one parameter, that is to say an operating or evaluation parameter, is adapted or selected as a function of the at least one control parameter, so that, for example, a characteristic curve or the like can be used.

For example, the operating and evaluation parameters of the sensor system can be adapted such that, depending on the control parameter, the sampling frequency of the sensor and / or an illumination range of the sensor system and / or the intensity of illumination and / or the detection range of the sensor and / or a number of objects to be evaluated and / or used in the evaluation filter and / or an evaluation accuracy can be adjusted. This will be explained in more detail with reference to some examples.

For example, a driving parameter can be determined on the basis of a light spectrum determined by a light sensor, for example an infrared light sensor. Because a sensor system of environmental sensors has to struggle with other challenges in the open air than, for example, within housings (eg parking garage, crash hall, etc.). With the aid of the light sensor, however, such different starting situations can be detected on the basis of the light spectrum and the sensor system can be better adapted to the respective relevant environment conditions by, for example, masking or otherwise evaluating known interfering side effects by means of corresponding filters already applied in the sensor system. Also can be provided, for example, active optical sensors to adjust the illumination range, for example, in the form of the amount of light. The function of the vehicle system is then improved adapted to the situation prepared input data and can be performed improved.

Similarly, a control parameter can be determined, for example, on the basis of the data of a rain sensor, for example, in environmental sensor systems to selectively hide rain effects and / or spray-generated ghost objects, for example in the case of ultrasonic sensors, already in the sensor system.

Other examples include, for example, a near vision and a far vision mode of operation of an environmental sensor system in a single function and the like.

In general, it can thus be formulated that it can be provided that a control parameter describing the current driving situation is used. In this way, it is thus possible to achieve an adaptation of the sensor characteristic in situ during the operation of a function, so that the sensor system delivers more suitable data in the current driving situation, ie in particular is optimally operated in accordance with the driving situation.

It should be noted at this point, moreover, that when sharing a sensor system by multiple functions possible conflicts can be resolved by a simple prioritization, so that, for example, safety functions receive a higher priority than comfort functions.

The method according to the invention can also be implemented particularly advantageously in collision-related functions of a predictive safety system. It can thus be provided that a parameter relating to a possible collision, in particular a parameter indicating the unavoidability of a collision, and / or a parameter indicating the exceeding of a threshold value by the collision probability is used as the control parameter. Just when a collision is inevitable is a particularly accurate determination of certain quantities, such as the time to collision or the collision speed, required to allow, for example, the triggering time of security systems and / or their parameterization in the safest possible way.

Conveniently, it can then be provided that in an operating mode which is active in the event of an unavoidable collision and / or when the threshold value is exceeded by the collision probability, the frequency in which the sensor system supplies data is increased, in particular doubled or quadrupled. Ultimately, if a collision can no longer be avoided, the data rate of the sensor system is increased overall, for example from 25 Hz to 50 or 100 Hz, which can be achieved by various concrete measures. Thus, it can be provided that in an operating mode which is active in the event of an unavoidable collision, the sensor is operated and / or read out at a higher sampling frequency and / or a number of evaluated sensed objects is reduced.

Firstly, the sensor itself, in the case of an active sensor system possibly also a transmitting device such as a lighting device, can deliver raw data in a higher frequency, which also leads to a higher data rate of the sensor system. Specifically, this means that operating parameters are adjusted, in the case of a solid-state detector, for example, the integration time is reduced or, in the case of a lighting device, for example an active infrared sensor, reduces or increases the length or frequency of the emitted light pulses.

However, it is also possible, in particular additionally, for the processing carried out by the evaluation module to be accelerated, for example by reducing a number of sensed objects evaluated in an operating mode which is active in an unavoidable collision and / or when the threshold value is exceeded by the collision probability. If detected objects are summarized in an object list, the maximum of the objects contained therein can be reduced, for example, from eight to four, so that z. B. only the collision object (as a CU object, "crash unavoidable "object) and three other, similarly close objects must be considered, thereby further increasing the frequency at which data is delivered.

It should also be noted at this point that even a reduction of the detection range (for example, by reducing the amount of light emitted by a light source) can favor a faster data frequency, since, for. B. by a reduced coverage area, less data must be processed.

Increasing the frequency at which the sensor system provides data provides a variety of benefits. On the one hand, the relative error caused by prediction operations of the function depends on the sampling rate, ie the frequency with which new data are available, since the prediction periods change. In particular, at low relative speeds of the motor vehicle and the collision object, for example, up to 25km / h relative speed, a very clear effect. This will be explained in more detail with reference to an example. If the relative speed in an imminent collision of a truck with a car is 15 km / h and if the premise for the prediction is maximum acceleration / braking, then a data frequency of 25 Hz results in an error range of approximately 16% with respect to the relative speed, which is square on the kinetic energy, which then has a tolerance range of about -18% to + 14%. At 100 Hz, the error range is only about 4%.

An increased data frequency / sampling rate also allows a more accurate determination of the time to collision (collision beginning). The higher number of measurements before the collision also allows a more accurate prediction of the relative velocity and the collision coverage. The higher validation quality allows a stronger influence on the tripping behavior. In particular, it should be pointed out that, of course, the error caused by measurement inaccuracy is also reduced by a larger number of measured values on the basis of statistics, in particular suitable statistical methods can be used.

Finally, the increased sampling rate can save resources, because a very complex and resource-intensive calculation method for the prediction of the collision parameters can be avoided, for example a trajectory calculation or the use of large buffers in the storage of history values.

In a further refinement of the method according to the invention, it is possible to operate the sensor system comprising an environmental sensor in a short-range mode in an operating mode which is active in the event of an unavoidable collision and / or when the threshold value is exceeded. Ultimately, when considering an imminent collision, only the near range is relevant, so that the data acquisition and processing can be optimized to that effect. Specifically, it can be provided that the sensor is operated with a smaller detection range in an operating mode that is active in the event of an unavoidable collision and / or when the threshold value is exceeded by the collision probability and / or an illumination range of the sensor or an illumination device assigned to the sensor is reduced and / or reduced for the near field optimized object filter, in particular a Kalman filter, is used for the evaluation and / or an object filter is adapted with respect to the near field.

Consequently, for example, for the remaining objects in the near range, the illumination in an active system or the detection range can be reduced, for example from 30 m to 10-15 m. By reducing the illumination range, for example, unpleasant side effects such as halo effects can be avoided, so that the data quality is improved.

Also by an adaptation of the object filter, for example a Kalman filter, especially for the near range, a further improvement of the data quality is possible if necessary. Such object filters are known and are used in particular in imaging sensor systems, if several consecutive Data Collections Properties of objects (represented as pixel groups, for example) should be determined. An example here is also a mean value filter. If these are optimized in the operating mode for the near range, then for example the object width, the object position and the object speed can be determined more accurately.

As already mentioned, usually different collision parameters are determined by the function, which results in various control commands, in particular for security systems, but possibly also driving interventions. For example, from data of the sensor system a collision time and / or at least one trigger time for a safety system of the motor vehicle can be determined.

Especially in the case of the predictive safety system results in a total of high benefits for the driver of the motor vehicle. The triggering strategy of the safety systems is optimized and the crash detection is improved. In addition, the repair costs are reduced when collisions with lower crash severity are achieved. It also results in a higher protection potential for heavy impact collisions and a reduced risk of injury by false triggering of the airbag in collisions with low crash severity. These are just a few of the resulting varied benefits.

In addition to the method, the invention also relates to a motor vehicle, comprising a sensor system and data of the sensor system using vehicle system, designed for carrying out the method according to the invention. All statements relating to the method can be analogously transferred to the motor vehicle according to the invention, so that even the aforementioned advantages are achieved. In particular, the vehicle system executing the function may be a predictive safety system, but the invention is fundamentally applicable to all the systems detecting the environment.

Fig. 1

ein erfindungsgemäßes Kraftfahrzeug, und

Fig. 2

einen Ablaufplan des erfindungsgemäßen Verfahrens.

Fig. 1 zeigt die Prinzipskizze eines erfindungsgemäßen Kraftfahrzeugs 1. Es umfasst neben anderen Fahrzeugsystemen, die bei 2 allgemein angedeutet sind, ein vorausschauendes Sicherheitssystem 3, welches als Funktion eine Kollisionsüberwachung aufweist, die bei drohender bzw. unvermeidbarer Kollision andere Sicherheitssysteme und Fahrzeugsysteme 2 des Kraftfahrzeugs 1, beispielsweise die Bremsen, Gurtsysteme, Airbags, eine anstellbare Motorhaube und dergleichen, ansteuert.Further advantages and details of the present invention will become apparent from the embodiments described below and with reference to the drawing. Showing:

Fig. 1

a motor vehicle according to the invention, and

Fig. 2

a flow chart of the method according to the invention.

Fig. 1 It comprises, in addition to other vehicle systems, which are generally indicated at 2, a predictive safety system 3, which has as a function of collision monitoring, which in the event of imminent or unavoidable collision other safety systems and vehicle systems 2 of the motor vehicle 1, for example the brakes, belt systems, airbags, a engagable hood and the like, controls.

In order to obtain information about the environment of the motor vehicle 1, the function receives input data from sensor systems, wherein a sensor system 4 is to be considered here in more detail. For communication with the sensor system 4 and the other vehicle systems, a communication bus, here a CAN bus 5, is provided.

The sensor system 4, here an environment sensor system, is an active optical sensor system 4, which comprises an infrared sensor 6 comprising a multiplicity of detector pixels and an illuminating device 7 which is directed in the same direction. About a function module 8, which is designed here as a processor, the raw data of the sensor 6 are processed and provided as object lists on the CAN bus 5 the vehicle systems 2, 3 available.

By means of its pixels, the infrared sensor 6 measures the distance (transit time measurement), the reflectivity, and ultimately also takes up a kind of image from which shapes of the objects can be deduced. By the sensor principle used in each case, additional information such. B. 3D capture, Reflection behavior and the like are available. An object list is then created by ultimately identifying groups of pixels as objects and looking at their timing, using various object filters to compensate for noise in the measurements and the like, such as Kalman filters. Thus, for each object, the object width, the object position and the object speed can be made available as prepared data in the object list.

Thus, there are some parameters for the operation of the sensor system 4, which determine the type and quality of the processed data. As an operating parameter of the sensor 6 (and the illuminating device 7, which is directly assigned to the sensor 6 here), for example, the amount of light to be emitted, the length of the light pulses, the frequency of the light pulses, the integration time for the detector pixels in the recording of the measured data and the sampling frequency , Of course, other fundamentally influenceable operating parameters are conceivable. But also the evaluation by the function module 8 basically has evaluation parameters, for example by the type of object filter used or other evaluation parameters.

The object list as processed data is now available via the CAN bus 5 as at least part of the input data of the "collision monitoring" function of the predictive safety system 3. The function finally checks the input data for a probability of collision and measures to be initiated and can be implemented as a software package or else comprehensively electronic components. Finally, input data are processed into operating parameters or control commands to other vehicle systems 2.

In normal operation of collision monitoring, all objects are also relevant in a wider range around the motor vehicle. Their relative speeds can be compared with the technical possibilities of the motor vehicle 1 (physical limit range) to a collision probability determine, as is well known and need not be explained here. Therefore, the sensor system 4 is usually configured to generate data for the wide-area, through a wide illumination range, many processed objects and a consequent rather low frequency, for example 25 Hz, with which the processed data is available to the vehicle systems 2, 3 be put.

The method according to the invention in this embodiment, however, provides two operating modes of the sensor system 4 to be used within the function of collision monitoring, namely a fast short-range mode and a slower long-range mode, which forms the standard setting as described above. In the present case, this is controlled on the basis of a drive parameter describing the unavoidability of a collision, which is determined internally in the function. Of course, other driving parameters can be taken into account, for example the time until the beginning of a collision.

Namely, in step 9, Fig. 2 Having determined that a collision is unavoidable, for example, will occur in 300-600 ms, so in a step 10, the sensor system 4 is transferred to the operation of the function collision monitoring in another operating mode, namely the short-range mode. For this purpose, the following operating and evaluation parameters of the sensor system 4 are changed. The processed data are optimized as a result only in higher frequency and especially in the near field, so more precisely, provided.

The evaluation of the object lists is reduced from, for example, greater than or equal to eight objects to four objects, in this case to the collision object (CU object) and further three objects with a similar distance to the motor vehicle 1. Thus, the framerate of the data transmission to 50 Hz can be at least doubled become.

For the remaining objects in the vicinity of interest, the illumination is reduced by the lighting device 7, for example from 30 meters to 10 to 15 meters. As a result, only the really relevant area is considered and unwanted side effects, such as halo effects, are significantly reduced. The data quality for the near range increases. In this way, the data frequency can be further increased, so that the object lists can be transmitted, for example, at 75 to 100 Hz.

It should be noted at this point that analogously, the integration times at the sensor 6 and its sampling frequency can be lowered or increased, which can cause a further increase in the data frequency.

Furthermore, an adaptation of the object filter is carried out specifically for the near field, so as to enable a further improvement of the data quality. In this way, the precision in determining object width, object position and object speed is increased. In fact, as already mentioned, the relative error is thereby significantly reduced.

In particular, the increase in the data frequency has a positive influence on the continued function of the function in step 11, which determines control commands, operating parameters and activation times for various other safety systems, based on intermediate parameters such as time to collision, collision coverage, collision speed and the like , These intermediate parameters can be more accurately determined in the near range mode of the sensor system 4. The higher number of measurements makes it possible to reduce inaccuracies through statistical methods.

Overall, therefore, results in a higher prediction quality and consequent improved control of other safety and vehicle systems. The crash severity can possibly be reduced and the protection potential is increased.

Claims (10)

Method for controlling the operation of at least one sensor system encompassing the surrounding area, comprising at least one sensor and / or at least one evaluation module of a motor vehicle, which sensor system supplies at least a part of the input data of an input data to concrete control actions of the vehicle evaluating function of a vehicle system,
characterized,
in that, depending on at least one control parameter determined in the context of the function, the sensor system is switched over between at least two operating modes used in the function. A method according to claim 1, characterized in that depending on the control parameters, the sampling frequency of the sensor and / or an illumination range of the sensor system and / or the intensity of illumination and / or the detection range of the sensor and / or a number of objects to be evaluated and / or in the frame the evaluation used filter and / or an evaluation accuracy can be adjusted. A method according to claim 1 or 2, characterized in that a current driving situation descriptive control parameters is used. Method according to one of the preceding claims, characterized in that a parameter relating to a possible collision, in particular a parameter indicating the unavoidability of a collision and / or a parameter indicating the exceeding of a threshold value by the collision probability, is used as the control parameter. Method according to Claim 4, characterized in that, in an operating mode which is active in the event of an unavoidable collision and / or when the threshold value is exceeded by the collision probability, the frequency in which the sensor system supplies data is increased, in particular doubled or quadrupled. A method according to claim 5, characterized in that operated in an unavoidable collision and / or exceeding the threshold value by the collision probability operating mode of the sensor with a higher sampling frequency and / or read and / or a number of evaluated sensed objects is reduced. Method according to one of Claims 4 to 6, characterized in that, in an operating mode which is active in the event of an unavoidable collision and / or when the threshold value is exceeded by the probability of collision, the sensor system comprising an environmental sensor is operated in a short-range mode. A method according to claim 7, characterized in that in an active in an unavoidable collision and / or exceeding the threshold by the collision probability operating mode, the sensor is operated with a smaller detection range and / or an illumination range of the sensor or a sensor associated with the illumination device is reduced and / or an object filter optimized for the near field, in particular a Kalman filter, is used for the evaluation and / or an object filter is adapted with regard to the near field. Method according to one of claims 4 to 8, characterized in that from data of the sensor system, a collision time and / or at least one triggering time for a safety system of the motor vehicle are determined. Motor vehicle (1) comprising a sensor system (4) and data of the sensor system (4) using the vehicle system (3), adapted for carrying out a method according to any one of the preceding claims.

Images